Skip to content
Snippets Groups Projects
Commit 4d705375 authored by Jakhes's avatar Jakhes
Browse files

Adding Kmeans to the Project

parent f35cbdc1
Branches
Tags
No related merge requests found
Show whitespace changes
Inline Side-by-side
src/helper_files/helper.cpp
+ 6
− 0
View file @ 4d705375
......@@ -44,6 +44,12 @@ float *convertToArray(vector<double> vec)
return convertToArray(newVec);
}
float *convertToArray(Row<size_t> vec)
{
colvec newVec = conv_to<colvec>::from(vec);
return convertToArray(newVec);
}
float *convertToArray(vector<vec> matrix)
{
vec newVec = matrix[0];
......
src/helper_files/helper.hpp
+ 2
− 0
View file @ 4d705375
......@@ -17,6 +17,8 @@ float *convertToArray(vector<size_t> vec);
float *convertToArray(vector<double> vec);
float *convertToArray(Row<size_t> vec);
float *convertToArray(vector<vec> vec);
......
src/methods/kmeans/Makefile 0 → 100644
+ 8
− 0
View file @ 4d705375
splfr=/usr/local/sicstus4.7.1/bin/splfr
METHOD_NAME=kmeans
$(METHOD_NAME).so: $(METHOD_NAME).pl $(METHOD_NAME).cpp
$(splfr) -larmadillo -fopenmp -lmlpack -lstdc++ -cxx --struct $(METHOD_NAME).pl $(METHOD_NAME).cpp ../../helper_files/helper.cpp
clean:
rm $(METHOD_NAME).so
src/methods/kmeans/kmeans.cpp 0 → 100644
+ 232
− 0
View file @ 4d705375
#include <sicstus/sicstus.h>
/* ex_glue.h is generated by splfr from the foreign/[2,3] facts.
Always include the glue header in your foreign resource code.
*/
#include "kmeans_glue.h"
#include <mlpack/methods/kmeans/kmeans.hpp>
// Include initialization strategies.
#include <mlpack/methods/kmeans/sample_initialization.hpp>
#include <mlpack/methods/kmeans/random_partition.hpp>
// Include empty cluster policies.
#include <mlpack/methods/kmeans/max_variance_new_cluster.hpp>
#include <mlpack/methods/kmeans/kill_empty_clusters.hpp>
#include <mlpack/methods/kmeans/allow_empty_clusters.hpp>
// Include Lloyd step types.
#include <mlpack/methods/kmeans/dual_tree_kmeans.hpp>
#include <mlpack/methods/kmeans/elkan_kmeans.hpp>
#include <mlpack/methods/kmeans/hamerly_kmeans.hpp>
#include <mlpack/methods/kmeans/pelleg_moore_kmeans.hpp>
#include <mlpack/core.hpp>
// including helper functions for converting between arma structures and arrays
#include "../../helper_files/helper.hpp"
// some of the most used namespaces
using namespace arma;
using namespace mlpack;
using namespace std;
using namespace mlpack::kmeans;
using namespace mlpack::metric;
// internal Funktion
// initiats KMeans with the given Class Template and inputs the params and returns the results
template< template<class,class> class LloydStepType=NaiveKMeans>
void InitAndClusterKMeans(SP_integer maxIterations, SP_integer initialPartition, SP_integer emptyCluster, float *dataMatArr, SP_integer dataMatSize, SP_integer dataMatRowNum, SP_integer clusters,
float **assignmentsArr, SP_integer *assignmentsArrSize, float **centroidsMatArr, SP_integer *centroidsMatColNum, SP_integer *centroidsMatRowNum)
{
// convert the Prolog arrays to arma::mat
mat data = convertArrayToMat(dataMatArr, dataMatSize, dataMatRowNum);
// get the ReturnVector
arma::Row<size_t> toReturnVector;
// get the ReturnMat
mat toReturnMat;
switch (initialPartition)
{
case 0:
// SampleInitialization
cout << "0" << endl;
switch (emptyCluster)
{
case 0:
// MaxVarianceNewCluster
cout << "0" << endl;
KMeans<EuclideanDistance, SampleInitialization, MaxVarianceNewCluster, LloydStepType>(maxIterations).Cluster(data, clusters, toReturnVector, toReturnMat, false, false);
break;
case 1:
// KillEmptyClusters
cout << "1" << endl;
KMeans<EuclideanDistance, SampleInitialization, KillEmptyClusters, LloydStepType>(maxIterations).Cluster(data, clusters, toReturnVector, toReturnMat, false, false);
break;
case 2:
// AllowEmptyClusters
cout << "2" << endl;
KMeans<EuclideanDistance, SampleInitialization, AllowEmptyClusters, LloydStepType>(maxIterations).Cluster(data, clusters, toReturnVector, toReturnMat, false, false);
break;
default:
// Wrong input
cout << "wrong input of emptyCluster, should be 0, 1 or 2" << endl;
break;
}
break;
case 1:
// RandomPartition
cout << "1" << endl;
switch (emptyCluster)
{
case 0:
// MaxVarianceNewCluster
cout << "0" << endl;
KMeans<EuclideanDistance, RandomPartition, MaxVarianceNewCluster, LloydStepType>(maxIterations).Cluster(data, clusters, toReturnVector, toReturnMat, false, false);
break;
case 1:
// KillEmptyClusters
cout << "1" << endl;
KMeans<EuclideanDistance, RandomPartition, KillEmptyClusters, LloydStepType>(maxIterations).Cluster(data, clusters, toReturnVector, toReturnMat, false, false);
break;
case 2:
// AllowEmptyClusters
cout << "2" << endl;
KMeans<EuclideanDistance, RandomPartition, AllowEmptyClusters, LloydStepType>(maxIterations).Cluster(data, clusters, toReturnVector, toReturnMat, false, false);
break;
default:
// Wrong input
cout << "wrong input of emptyCluster, should be 0, 1 or 2" << endl;
break;
}
break;
default:
// Wrong input
cout << "wrong input of initialPartition, should be 0 or 1" << endl;
break;
}
// return the Vector lenght
*assignmentsArrSize = toReturnVector.n_elem;
// return the Vector as Array
*assignmentsArr = convertToArray(toReturnVector);
// return the Matrix dimensions
*centroidsMatColNum = toReturnMat.n_cols;
*centroidsMatRowNum = toReturnMat.n_rows;
// return the Matrix as one long Array
*centroidsMatArr = convertToArray(toReturnMat);
}
// TODO:
// input: const size_t maxIterations = 1000,
// const InitialPartitionPolicy partitioner = InitialPartitionPolicy(),
// const EmptyClusterPolicy emptyClusterAction = EmptyClusterPolicy()
//
// const MatType & data,
// const size_t clusters,
// arma::Row< size_t > & assignments,
// arma::mat & centroids,
// const bool initialAssignmentGuess = false,
// const bool initialCentroidGuess = false
// output:
// description:
void naiveKMeans(SP_integer maxIterations, SP_integer initialPartition, SP_integer emptyCluster, float *dataMatArr, SP_integer dataMatSize, SP_integer dataMatRowNum, SP_integer clusters,
float **assignmentsArr, SP_integer *assignmentsArrSize, float **centroidsMatArr, SP_integer *centroidsMatColNum, SP_integer *centroidsMatRowNum)
{
InitAndClusterKMeans<NaiveKMeans>(maxIterations, initialPartition, emptyCluster, dataMatArr, dataMatSize, dataMatRowNum, clusters,
assignmentsArr, assignmentsArrSize, centroidsMatArr, centroidsMatColNum, centroidsMatRowNum);
}
// TODO:
// input: const size_t maxIterations = 1000,
// const InitialPartitionPolicy partitioner = InitialPartitionPolicy(),
// const EmptyClusterPolicy emptyClusterAction = EmptyClusterPolicy()
//
// const MatType & data,
// const size_t clusters,
// arma::Row< size_t > & assignments,
// arma::mat & centroids,
// const bool initialAssignmentGuess = false,
// const bool initialCentroidGuess = false
// output:
// description:
void dualTreeKMeans(SP_integer maxIterations, SP_integer initialPartition, SP_integer emptyCluster, float *dataMatArr, SP_integer dataMatSize, SP_integer dataMatRowNum, SP_integer clusters,
float **assignmentsArr, SP_integer *assignmentsArrSize, float **centroidsMatArr, SP_integer *centroidsMatColNum, SP_integer *centroidsMatRowNum)
{
InitAndClusterKMeans<DualTreeKMeans>(maxIterations, initialPartition, emptyCluster, dataMatArr, dataMatSize, dataMatRowNum, clusters,
assignmentsArr, assignmentsArrSize, centroidsMatArr, centroidsMatColNum, centroidsMatRowNum);
}
// TODO:
// input: const size_t maxIterations = 1000,
// const InitialPartitionPolicy partitioner = InitialPartitionPolicy(),
// const EmptyClusterPolicy emptyClusterAction = EmptyClusterPolicy()
//
// const MatType & data,
// const size_t clusters,
// arma::Row< size_t > & assignments,
// arma::mat & centroids,
// const bool initialAssignmentGuess = false,
// const bool initialCentroidGuess = false
// output:
// description:
void elkanKMeans(SP_integer maxIterations, SP_integer initialPartition, SP_integer emptyCluster, float *dataMatArr, SP_integer dataMatSize, SP_integer dataMatRowNum, SP_integer clusters,
float **assignmentsArr, SP_integer *assignmentsArrSize, float **centroidsMatArr, SP_integer *centroidsMatColNum, SP_integer *centroidsMatRowNum)
{
InitAndClusterKMeans<ElkanKMeans>(maxIterations, initialPartition, emptyCluster, dataMatArr, dataMatSize, dataMatRowNum, clusters,
assignmentsArr, assignmentsArrSize, centroidsMatArr, centroidsMatColNum, centroidsMatRowNum);
}
// TODO:
// input: const size_t maxIterations = 1000,
// const InitialPartitionPolicy partitioner = InitialPartitionPolicy(),
// const EmptyClusterPolicy emptyClusterAction = EmptyClusterPolicy()
//
// const MatType & data,
// const size_t clusters,
// arma::Row< size_t > & assignments,
// arma::mat & centroids,
// const bool initialAssignmentGuess = false,
// const bool initialCentroidGuess = false
// output:
// description:
void hamerlyKMeans(SP_integer maxIterations, SP_integer initialPartition, SP_integer emptyCluster, float *dataMatArr, SP_integer dataMatSize, SP_integer dataMatRowNum, SP_integer clusters,
float **assignmentsArr, SP_integer *assignmentsArrSize, float **centroidsMatArr, SP_integer *centroidsMatColNum, SP_integer *centroidsMatRowNum)
{
InitAndClusterKMeans<HamerlyKMeans>(maxIterations, initialPartition, emptyCluster, dataMatArr, dataMatSize, dataMatRowNum, clusters,
assignmentsArr, assignmentsArrSize, centroidsMatArr, centroidsMatColNum, centroidsMatRowNum);
}
// TODO:
// input: const size_t maxIterations = 1000,
// const InitialPartitionPolicy partitioner = InitialPartitionPolicy(),
// const EmptyClusterPolicy emptyClusterAction = EmptyClusterPolicy()
//
// const MatType & data,
// const size_t clusters,
// arma::Row< size_t > & assignments,
// arma::mat & centroids,
// const bool initialAssignmentGuess = false,
// const bool initialCentroidGuess = false
// output:
// description:
void pellegMooreKMeans(SP_integer maxIterations, SP_integer initialPartition, SP_integer emptyCluster, float *dataMatArr, SP_integer dataMatSize, SP_integer dataMatRowNum, SP_integer clusters,
float **assignmentsArr, SP_integer *assignmentsArrSize, float **centroidsMatArr, SP_integer *centroidsMatColNum, SP_integer *centroidsMatRowNum)
{
InitAndClusterKMeans<PellegMooreKMeans>(maxIterations, initialPartition, emptyCluster, dataMatArr, dataMatSize, dataMatRowNum, clusters,
assignmentsArr, assignmentsArrSize, centroidsMatArr, centroidsMatColNum, centroidsMatRowNum);
}
\ No newline at end of file
src/methods/kmeans/kmeans.pl 0 → 100644
+ 79
− 0
View file @ 4d705375
:- module(kmeans, [ naiveKMeans/12,
dualTreeKMeans/12,
elkanKMeans/12,
hamerlyKMeans/12,
pellegMooreKMeans/12]).
%% requirements of library(struct)
:- load_files(library(str_decl),
[when(compile_time), if(changed)]).
%% needed for using the array type
:- use_module(library(structs)).
:- use_module('../../helper_files/helper.pl').
%% type definitions for the float array
:- foreign_type
float32 = float_32,
float_array = array(float32).
%% definitions for the connected function
%% TODO:
%% input:
%% output:
%% description:
foreign(naiveKMeans, c, naiveKMeans(+integer, +integer, +integer, +pointer(float_array), +integer, +integer, +integer,
-pointer(float_array), -integer, -pointer(float_array), -integer, -integer)).
%% TODO:
%% input:
%% output:
%% description:
foreign(dualTreeKMeans, c, dualTreeKMeans(+integer, +integer, +integer, +pointer(float_array), +integer, +integer, +integer,
-pointer(float_array), -integer, -pointer(float_array), -integer, -integer)).
%% TODO:
%% input:
%% output:
%% description:
foreign(elkanKMeans, c, elkanKMeans(+integer, +integer, +integer, +pointer(float_array), +integer, +integer, +integer,
-pointer(float_array), -integer, -pointer(float_array), -integer, -integer)).
%% TODO:
%% input:
%% output:
%% description:
foreign(hamerlyKMeans, c, hamerlyKMeans(+integer, +integer, +integer, +pointer(float_array), +integer, +integer, +integer,
-pointer(float_array), -integer, -pointer(float_array), -integer, -integer)).
%% TODO:
%% input:
%% output:
%% description:
foreign(pellegMooreKMeans, c, pellegMooreKMeans(+integer, +integer, +integer, +pointer(float_array), +integer, +integer, +integer,
-pointer(float_array), -integer, -pointer(float_array), -integer, -integer)).
%% +integer , +float32
%% [-integer] , [-float32]
%% matrix input
%% +pointer(float_array), +integer, +integer
%% array input
%% +pointer(float_array), +integer
%% matrix return
%% -pointer(float_array), -integer, -integer
%% array return
%% -pointer(float_array), -integer
%% Defines the functions that get connected from main.cpp
foreign_resource(kmeans, [ naiveKMeans,
dualTreeKMeans,
elkanKMeans,
hamerlyKMeans,
pellegMooreKMeans]).
:- load_foreign_resource(kmeans).
\ No newline at end of file
src/methods/kmeans/kmeans_test.pl 0 → 100644
+ 60
− 0
View file @ 4d705375
:- use_module(library(plunit)).
:- use_module(kmeans).
:- use_module('../../helper_files/helper.pl').
:- begin_tests(lists).
%% alpha tests
test(naiveKMeans) :-
convert_list_to_float_array([5.1,3.5,1.4,4.9,3.0,1.4,4.7,3.2,1.3,4.6,3.1,1.5, 1.4, 3.0, 2.1, 0.1],4, array(Xsize, Xrownum, X)),
naiveKMeans(20, 0, 1, X, Xsize, Xrownum, 3, Y, Ysize, Centroids, Centroidscolnum, Centroidsrownum),
print(Ysize),
convert_float_array_to_list(Y, Ysize, ResultsY),
print(ResultsY),
print(Centroidscolnum),
convert_float_array_to_2d_list(Centroids, Centroidscolnum, Centroidsrownum, Results),
print(Results).
test(dualTreeKMeans) :-
convert_list_to_float_array([5.1,3.5,1.4,4.9,3.0,1.4,4.7,3.2,1.3,4.6,3.1,1.5, 1.4, 3.0, 2.1, 0.1],4, array(Xsize, Xrownum, X)),
dualTreeKMeans(20, 0, 1, X, Xsize, Xrownum, 3, Y, Ysize, Centroids, Centroidscolnum, Centroidsrownum),
print(Ysize),
convert_float_array_to_list(Y, Ysize, ResultsY),
print(ResultsY),
print(Centroidscolnum),
convert_float_array_to_2d_list(Centroids, Centroidscolnum, Centroidsrownum, Results),
print(Results).
test(elkanKMeans) :-
convert_list_to_float_array([5.1,3.5,1.4,4.9,3.0,1.4,4.7,3.2,1.3,4.6,3.1,1.5, 1.4, 3.0, 2.1, 0.1],4, array(Xsize, Xrownum, X)),
elkanKMeans(20, 0, 1, X, Xsize, Xrownum, 3, Y, Ysize, Centroids, Centroidscolnum, Centroidsrownum),
print(Ysize),
convert_float_array_to_list(Y, Ysize, ResultsY),
print(ResultsY),
print(Centroidscolnum),
convert_float_array_to_2d_list(Centroids, Centroidscolnum, Centroidsrownum, Results),
print(Results).
test(hamerlyKMeans) :-
convert_list_to_float_array([5.1,3.5,1.4,4.9,3.0,1.4,4.7,3.2,1.3,4.6,3.1,1.5, 1.4, 3.0, 2.1, 0.1],4, array(Xsize, Xrownum, X)),
hamerlyKMeans(20, 0, 1, X, Xsize, Xrownum, 3, Y, Ysize, Centroids, Centroidscolnum, Centroidsrownum),
print(Ysize),
convert_float_array_to_list(Y, Ysize, ResultsY),
print(ResultsY),
print(Centroidscolnum),
convert_float_array_to_2d_list(Centroids, Centroidscolnum, Centroidsrownum, Results),
print(Results).
test(pellegMooreKMeans) :-
convert_list_to_float_array([5.1,3.5,1.4,4.9,3.0,1.4,4.7,3.2,1.3,4.6,3.1,1.5, 1.4, 3.0, 2.1, 0.1],4, array(Xsize, Xrownum, X)),
pellegMooreKMeans(20, 0, 0, X, Xsize, Xrownum, 3, Y, Ysize, Centroids, Centroidscolnum, Centroidsrownum),
print(Ysize),
convert_float_array_to_list(Y, Ysize, ResultsY),
print(ResultsY),
print(Centroidscolnum),
convert_float_array_to_2d_list(Centroids, Centroidscolnum, Centroidsrownum, Results),
print(Results).
:- end_tests(lists).
\ No newline at end of file
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment